prioritize
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external project artifacts to perform ranking calculations, establishing a surface for indirect prompt injection.
- Ingestion points: Reads planned work items from the
tasks/directory, user-defined file paths, or inline text descriptions as specified in Phase 0 and Workflow Step 1. - Boundary markers: The instructions lack explicit delimiters or safety prompts to prevent the agent from following instructions that might be embedded within the project tasks themselves.
- Capability inventory: The skill is capable of writing to the file system (modifying task frontmatter) and executing tool commands like
$task-trackand theask user question toolfor workflow escalation. - Sanitization: The skill does not implement sanitization or validation of the ingested task content before using the model to analyze and score it.
Audit Metadata